Marketeer™Loading…
Proof
Every figure on this page was computed by a query when you loaded it. None of them is typed into the HTML. Each one says what it counts, over what window, and as at when. Where a query did not answer, it says so rather than showing you the last number we happened to know. A zero is printed as a zero.
A figure that was measured shows the figure. A query that answered with no record says nothing is recorded yet. A query that did not answer says it could not be computed. A number we could compute but may not publish says it is withheld, with the gate that stops it.
Every exam and benchmark question is answered against an invented business. We are a processor of our customers’ content, and nothing on this page is derived from a customer’s workspace, aggregated or otherwise.
A platform’s attribution is never labelled as the cause of a sale. Too few enquiries to call a winner means collect more evidence, not invent one. A time saving is reported after counting briefing, reviewing and correcting, not before.
Kate is tested against a fixed exam and a fixed benchmark, both against an invented Leeds bakery, and a separate model, deliberately not the one being tested, marks every answer. We wrote the questions, so this is a self-set exam and we say so. The latest results are below, each with its date. When a stand-in model sits the exam, its score is not hers, so it is not published as hers.
Consultant exam pass rate
94% 17 of 53 answers scored
Latest result on record: 13 Sep 2026.
Latest result: 13 Sep 2026. 17 of 53 questions in the frozen panel produced an answer the judge could score, and the rate is over those alone: a question with no scorable answer is left out, not counted as a fail. A pass means a separate model (not the one being tested) rated the answer worth paying a consultant for. Sat by claude-opus-4-8 from Anthropic. Once the app's daily Opus limit is reached, answers come from a lower-cost fallback model instead, and a run does not record which of its answers did.
Exam runs kept on record
52
One stored row per run. First 12 Jul 2026, latest 13 Sep 2026.
Benchmark mean score
8.7 / 10 50 of 50 answers scored
Latest result on record: 9 Aug 2026.
Latest result: 9 Aug 2026. Mean of the judge's 0-10 scores across 50 of 50 questions, tiered easy to hard, all answered against one invented bakery. The model that sat this run was not recorded.
MarketingBench pass rate, public tasks
100%
Latest result: 4 Oct 2026. 9 of 9 scored public tasks passed, out of 13 published tasks. 3 further tasks are held back so the public list cannot be overfit, and are excluded from this figure.
Questions in the panels the scores above come from
123
60 in the frozen exam, 50 in the benchmark, and 13 published MarketingBench tasks you can read in full at /marketingbench. Every question is written by us and answered against an invented business, never a customer's own data. No human consultant was scored alongside.
An AI marketing department that only works when someone opens a tab is not a hire, it is a tool. So here is our own machine's attendance record: every scheduled job this deployment declares, judged against its own tolerated silence, computed when you loaded this page. A job that has never run and has not been watched long enough to judge is counted separately, as neither running nor broken, because alarming on ignorance is how a signal gets ignored.
Scheduled jobs registered
40
Every scheduled job this deployment declares, none excluded. A job cannot be added, removed or re-timed in vercel.json without this registry moving with it, because a test compares the two lists both ways.
Jobs that completed a run inside their own window
36 of 40
Counted from the completion stamp each job writes as its last statement, so a job that starts and dies does not count here, and neither does one that skipped all of its work.
Jobs not running
0
0 invoked and dying part-way through; 0 not being invoked at all. Which jobs, and what stops happening while they are quiet, is on our internal health endpoint rather than here.
Everything above is about us: an exam Kate is tested against, on an invented bakery, and our own machine’s attendance record. Nothing above is derived from a customer’s workspace, because we are a processor of that content and the purposes we published do not include making public statistics out of it. Anonymising it would itself be processing, so there is no aggregate small enough to slip through the gate. You will not find a “businesses served” counter here either. A headcount is not evidence that the work gets done, and when we do publish one it will be on this page with its query beside it, like everything else.
Signed permissions on file to publish a named customer's numbers
0
None. So there are no case studies on this site, and /customers is empty rather than populated with invented ones. When this moves, the case study appears and this figure moves with it.
Minimum businesses in any cell before we would publish a breakdown
20
Across the 8 business types we offer, and worse once you cross them with region, this threshold suppresses every cell we could build today. We do not compute the cell sizes: the lawful-basis gate below fails first, and running the query anyway would itself be the processing we say we are not doing.
Posts published for customers, by channel, over the last 30 days
Suppressed
Our agreement lets us use your content only to do your marketing, not to make public statistics out of it.
Reload the page. The generation stamp moves, and so does anything computed from the clock. Nothing here is baked into a build.
Open “How this is computed” under each figure. The source line names the row and the field the number came from, not a department.
The same ledger is served as a machine-readable file, never cached, so you can compare two loads.
The tasks and the rubric behind the MarketingBench figure are published in full at MarketingBench, including the ones we lose.
When a query fails, the figure disappears and says so. A number here with no stamp against it is a bug, and it is the only kind of bug this page cannot survive.
Real results go up when a customer agrees to share them, and not before. Until then the customers page says so.
The signed-permissions count above is the gate. What every case study will have to show is on the customers page.
Jobs that skipped their work
4
Invoked on schedule, then stopped before doing any of its work because the job is switched off or something it needs, such as a model key, is not configured. Counted here, not as running. A job that skips only part of its work still counts as running: with model tests switched off, MarketingBench still scores its rule-checked tasks and skips only the ones a model has to answer.
Jobs we cannot yet judge
0
Never stamped, and we have not been watching for longer than the job's own interval, so silence is not yet evidence. Reported as its own figure rather than counted as running or as broken.
We hold this and could total it in one query. We are a processor of workspace content and our published purposes are exhaustive: authenticating you, generating drafts, routing approvals, producing reports, supporting you. Deriving a public statistic is a new purpose, and anonymising it is itself processing, so an aggregate does not escape the gate. The fix is a contract change we have not made, not a bigger sample.
Businesses using Marketeer, broken down by trade and region
Suppressed
At our size, a trade and a town together would point to one real business, so we do not publish it.
Trade plus region plus a date is a quasi-identifier at any size, and at ours it is a name. Even if the basis existed, every cell would fall under the minimum above and be suppressed, which is why you would be reading a table of blanks rather than a benchmark.
Measured reach or engagement uplift after one of Kate's recommendations
Suppressed
These are each business's own sales and results. They belong to that business, not to our marketing.
These figures are the business's own trading performance and are confidential to them under our own terms. They are also platform-derived, and what we may do with platform insights beyond serving the customer who provided them is not ours to decide.
A named business, with the numbers we measured together
Suppressed
A business can be named here only after it has signed its agreement to be named.
This is the one place consent is the right instrument, because it is one named business agreeing to one named story. The wording is written and waiting. Until a signed one is on file, /customers stays empty. An empty page is not a weakness; a page of invented case studies would be.